Week 7: Logistic Regression
Logistic Regression
-Predictive analysis used to describe data and explain relationship
between one dependent binary variable and one or more
nominal/ordinal/interval/ratio level variables
- Binary classification (Y=1 true, Y=0 false)
-Binary categorical variables
Logit
- Goal = find function of predictor variables that relates them to a 0/1
outcome
- Logit modeled as linear function of predictors
- Mapped back to probability (p) and mapped to a class (0 or 1)
-P = probability for outcome =1
-
-
-dependent variable = loan, independent variable = income
- EX:
-
Classification analysis
-Determining the probability of an observation to be part of a certain
class or not
- Express probability with value between 0 and 1
-Confusion matrix = primary source of accuracy estimation
- Test data for which the true outcomes are known
-
-
-
- Training and validation/testing = split the data into two mutually
exclusive sets, to investigate the strength of the developed model
-Odds ratio = p / p-1
-
-
-Logistic = Bigger BIC is better
-Don’t look at r-square
-Multiple logistic = multiple independent variables
-EX: want to find body weight, calorie intake, fat intake and age
influence on probability of having heart attack (yes or no)
-Simple logistic = one independent variable
Powered by TCPDF (www.tcpdf.org)
